Table of ContentsAcknowledgement1. Introduction2. Ideal Problem Solving Capabilities3. Ideal Software Development Capabilities and Software Reuse as a Practical Approximation4. Integration of Development Capabilities5. Generic Model Improvement6. Features Analysis and Initial Formulation of the Generic Development and Application Model7. Conclusion and RecommendationReferencesTechnical ReportCMU/SEI-88-TR-022ESD-TR-88-023Perspectiveon Software ReuseJ. M. PerryGTE Resident AffiliateGTE Government Systems CorporationSeptember 1988Technical ReportCMU/SEI-88-TR-022ESD-TR-88-023September 1988Perspective on Software ReuseABJ. M. PerryGTE Resident AffiliateGTE Government Systems CorporationApplication of Reusable SoftwareComponents ProjectApproved for public release.Distribution unlimited.Software Engineering InstituteCarnegie Mellon UniversityPittsburgh, Pennsylvania 15213This technical report was prepared for theSEI Joint Program OfficeESD/XRSHanscom AFB, MA 01731The ideas and findings in this report should not be construed as an officialDoD position. It is published in the interest of scientific and technicalinformation exchange.Review and ApprovalThis report has been reviewed and is approved for publication.FOR THE COMMANDERKarl H. Shingler SIGNATURE ON FILESEI Joint Program OfficeThis work is sponsored by the U.S. Department of Defense.Copyright  1988 by Carnegie Mellon University.This document is available through the Defense Technical Information Center. DTIC provides access to and transfer ofscientific and technical information for DoD personnel, DoD contractors and potential contractors, and other U.S. Governmentagency personnel and their contractors. To obtain a copy, please contact DTIC directly: Defense Technical InformationCenter, Attn: FDRA, Cameron Station, Alexandria, VA 22304-6145.Copies of this document are also available through the National Technical Information Services. For information on ordering,please contact NTIS directly: National Technical Information Services, U.S. Department of Commerce, Springfield, VA 22161.Use of any trademarks in this handbook is not intended in any way to infringe on the rights of the trademark holder.AcknowledgementGTE, the SEI, and the Application of Reusable Components Project provided the oppor-tunity to formulate and refine the ideas of this paper.The report was edited by Linda Hutz Pesante who, both as a technical writer and objectivereader, contributed to the expression of these ideas and the completion of the report.Perspective on Software ReuseAbstract: This report presents a perspective on software reuse in the context of"ideal" software development capabilities. Software reuse is viewed as a meansof achieving—or at least approximating—the ideal capabilities. A generic appli-cation and development model is proposed for unifying various types of softwarereuse. The model can be initially formulated as a project family architecture andproduced from a domain features analysis. The approach presented in this reportis intended to lead to a reuse strategy and methodology for software development.1. IntroductionThis report focuses on an approach to software reuse which can be expanded to a reusestrategy for software development. We assume a methodology which divides the life cycleinto phases, such as requirements through integration and maintenance, where each phaseproduces intermediate artifacts.The report views software development as a form of means-ends analysis, a problem-solving method that formulates a problem as a search for a sequence of actions beginningwith an initial situation or state and ending with a goal situation or state. At each state, anoperator is selected which reduces the difference between the current state and the goal. Asolution to a problem is a sequence of operators, the sequence of states (from initial throughintermediate to the goal), and the constraints which the sequence satisfies. In this setting,the initial state is the beginning of a software development, that is, the software require-ments and system description; a goal is the software system product; and the operators aremethods for transforming the artifacts. Software engineering attempts to reduce the searchaspects of the process by formulating methodologies which specify the type of the interme-diate artifacts and guide the application of operators.In means-end analysis, analogical problem solving is the process of retrieving a solution to asimilar, previously solved problem and transforming it into a solution of the current problem.Using this approach, software reuse can be precisely formulated as analogical development,where a previous development from similar requirements is transformed to a new devel-opment satisfying new constraints. An important point is that the selection of operators cor-responds to design decisions. This view gives us insight into two reasons systematic soft-ware reuse is difficult. First, the complete solution to the previous software development istypically not available; usually, only the artifacts are available, and design decision infor-mation is incomplete. Secondly, the constraints of the solution to the current developmentmay differ greatly from the constraints of the previous development.This view of software reuse as the use of "sequences" of software development "solutions"of previous projects to "solve" a current development underlies many of the ideas of thefollowing discussion.CMU/SEI-88-TR-22 12 CMU/SEI-88-TR-222. Ideal Problem Solving CapabilitiesTo put software reuse in perspective, this section describes "ideal" capabilities for problemsolving; the next section reformulates them for software development. Finally, reuse ispresented as a way of achieving (or, at least, of approximating) these desired "ideal" capa-bilities. The capability is "ideal" in the sense of being powerful and efficient for certain appli-cation domains, and desirable but not yet available for other application domains. The"ideal" capabilities are presented as problem-solving scenarios. In the next section, they arerecast as software development scenarios so that they can become the basis for a reusestrategy.Problem solving, including software development, is a goal-directed process which attemptsto produce products (i.e., goal artifacts) from initial artifacts, such as requirements or specifi-cations. The development involves, to varying degrees, both constructive and derivativeprocesses. For